The development of effective cancer therapies has been a major focus of biomedical research. Surgical procedures have been developed and used to treat patients whose tumors are confined to particular anatomical sites. However, at present, only about 25% of patients have tumors that are truly confined and amenable to surgical treatment alone (Slapak et al. in Harrison's Principles of Internal Medicine (Isselbacher et al., eds.) McGraw-Hill, Inc., NY (1994) pp. 1826–1850). Radiation therapy, like surgery, is a local modality whose usefulness in the treatment of cancer depends to a large extent on the inherent radiosensitivity of the tumor and its adjacent normal tissues. However, radiation therapy is associated with both acute toxicity and long term sequelae. Furthermore, radiation therapy is known to be mutagenic, carcinogenic, and teratogenic (Slapak et al., ibid.).
Systemic chemotherapy alone or in combination with surgery and/or radiation therapy is currently the primary treatment available for disseminated malignancies. However, conventional chemotherapeutic agents which either block enzymatic pathways or randomly interact with DNA irrespective of the cell phenotype, lack specificity for killing neoplastic cells. Thus, systemic toxicity often results from standard cytotoxic chemotherapy. More recently, the development of agents that block replication, transcription, or translation in transformed cells, and at the same time defeat the ability of cells to become resistant, has been the goal of many approaches to chemotherapy.
One strategy is to down regulate the expression of a gene associated with the neoplastic phenotype in a cell. A technique for turning off a single activated gene is the use of antisense oligodeoxynucleotides and their analogues for inhibition of gene expression (Zamecnik et al. (1978) Proc. Natl. Acad. Sci. (USA) 75:280–284). An antisense oligonucleotide targeted at a gene involved in the neoplastic cell growth should specifically interfere only with the expression of that gene, resulting in arrest of cancer cell growth. The ability to specifically block or down-regulate expression of such genes provides a powerful tool to explore the molecular basis of normal growth regulation, as well as the opportunity for therapeutic intervention (see, e.g., Cho-Chung (1993) Curr. Opin. Thera. Patents 3:1737–1750). The identification of genes that confer a growth advantage to neoplastic cells as well as other genes causally related to cancer and the understanding of the genetic mechanism(s) responsible for their activation makes the antisense approach to cancer treatment possible.
One such gene encodes the RIα subunit of cyclic AMP (cAMP)-dependent protein kinase A (PKA) (Krebs (1972) Curr. Topics Cell. Regul. 5:99–133). Protein kinase is bound by cAMP, which is thought to have a role in the control of cell proliferation and differentiation (see, e.g., Cho-Chung (1980) J. Cyclic Nucleotide Res. 6:163–167) There are two types of PKA, type I (PKA-I) and type II (PKA-II), both of which share a common C subunit but each containing distinct R subunits, RI and Rh, respectively (Beebe et al. in The Enzymes: Control by Phosphorylation, 17(A) :43–111 (Academic, New York, 1986) . The R subunit isoforms differ in tissue distribution (Øyen et al. (1988) FEBS Lett. 229:391–394; Clegg et al. (1988) Proc. Natl. Acad. Sci. (USA) 85:3703–3707) and in biochemical properties (Beebe et al. in The Enzymes: Control by Phosphorylation, 17(A) : 43–111 (Academic Press, NY, 1986); Cadd et al. (1990) J. Biol. Chem. 265:19502–19506) . The two general isoforms of the R subunit also differ in their subcellular localization: RI is found throughout the cytoplasm; whereas RII localizes to nuclei, nucleoli, Golgi apparatus and the microtubule-organizing center (see, e.g., Lohmann in Advances in Cyclic Nucleotide and Protein Phosphorylation Research , 18: 63–117 (Raven, New York, 1984; and Nigg et al. (1985) Cell 41:1039–1051).
An increase in the level of RIα expression has been demonstrated in human cancer cell lines and in primary tumors, as compared with normal counterparts, in cells after transformation with the Ki-ras oncogene or transforming growth factor-α, and upon stimulation of cell growth with granulocyte-macrophage colony-stimulating factor (GM-CSF) or phorbol esters (Lohmann in Advances in Cyclic Nucleotide and Protein Phosphorylation Research, 18:63–117 (Raven, N.Y., 1984); and Cho-Chung (1990) Cancer Res. 50:7093–7100). Conversely, a decrease in the expression of RIα has been correlated with growth inhibition induced by site-selective cAMP analogs in a broad spectrum of human cancer cell lines (Cho-Chung (1990) Cancer Res. 50:7093–7100). It has also been determined that the expression of RI/PKA-I and RII/PKA-II has an inverse relationship during ontogenic development and cell differentiation (Lohmann in Advances in Cyclic Nucleotide and Protein Phosphorylation Research, Vol. 18, 63–117 (Raven, N.Y., 1984); Cho-Chung (1990) Cancer Res. 50:7093–7100). The RIα subunit of PKA has thus been hypothesized to be an ontogenic growth-inducing protein whose constitutive expression disrupts normal ontogenic processes, resulting in a pathogenic outgrowth, such as malignancy (Nesterova et al. (1995) Nature Medicine 1:528–533).
Antisense oligonucleotides directed to the RIα gene have been prepared. U.S. Pat. No. 5,271,941 describes phosphodiester-linked antisense oligonucleotides complementary to a region of the first 100 N-terminal amino acids of RIα which inhibit the expression of RIα in leukemia cells in vitro. In addition, antisense phosphorothioate oligodeoxynucleotides corresponding to the N-terminal 8–13 codons of the RIα gene was found to reduce in vivo tumor growth in nude mice (Nesterova et al. (1995) Nature Med. 1:528–533).
Unfortunately, problems have been encountered with the use of phosphodiester-linked (PO) oligonucleotides and some phosphorothioate-linked (PS) oligonucleotides. It is known that nucleases in the serum readily degrade PO oligonucleotides. Replacement of the phosphodiester internucleotide linkages with phosphorothioate internucleotide linkages has been shown to stabilize oligonucleotides in cells, cell extracts, serum, and other nuclease-containing solutions (see, e.g., Bacon et al. (1990) Biochem. Biophys. Meth. 20:259) as well as in vivo (Iversen (1993) Antisense Research and Application (Crooke, ed) CRC Press, 461). However, some PS oligonucleotides have been found to exhibit an immunostimulatory response, which in certain cases may be undesirable. For example, Galbraith et al. (Antisense Res. & Dev. (1994) 4:201–206) disclose complement activation by some PS oligonucleotides. Henry et al. (Pharm. Res. (1994) 11: PPDM8082) disclose that some PS oligonucleotides may potentially interfere with blood clotting.
There is, therefore, a need for modified oligonucleotides directed to cancer-related genes that retain gene expression inhibition properties while producing fewer side effects than conventional oligonucleotides.